Post

The method (and assumptions) are also described a bit: if you have any suggestions/ideas to improve the data I'd love to hear it!

0 reposts4 likes